• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) ¼¨Çø® °ª ÃßÁ¤À» ÀÌ¿ëÇÑ ÁÖ°¡ º¯È­ ¿øÀÎ ¼³¸í ¹®Àå ºÐ·ù ¸ðµ¨ ºÐ¼®
¿µ¹®Á¦¸ñ(English Title) Model Analysis using Estimation of Shapley Value on Classification of Sentences Explaining Causes of Changes in Stock Prices
ÀúÀÚ(Author) õ¿¹Àº   ¹Ú¿µÁø   ¼º³«¿ø   ÃÖÀç½Ä   Ye Eun Chun   Youngjin Park   Nakwon Sung   Jaesik Choi  
¿ø¹®¼ö·Ïó(Citation) VOL 26 NO. 04 PP. 0195 ~ 0201 (2020. 04)
Çѱ۳»¿ë
(Korean Abstract)
ÃÖ±Ù ÀΰøÁö´É ±â¼úÀÇ ¹ßÀüÀ¸·Î ÅõÀÚ ÀÚµ¿È­°¡ ÀÌ·ç¾îÁö°í ÀÖÁö¸¸, ÅõÀÚÀÚ°¡ ÀΰøÁö´ÉÀÇ ÁÖ°¡¿¹Ãø °á°ú¸¦ ÀÌÇØÇÏ°í ÆÇ´Ü ³»¸®±â´Â ´õ¿í ¾î·Á¿öÁ³´Ù. µû¶ó¼­ ÅõÀÚÀÚ°¡ ÀÌÇØÇÒ ¼ö ÀÖµµ·Ï ÁÖ°¡¿¹Ãø¿¡ ´ëÇÑ ¿øÀÎ ¼³¸í Á¦°øÀÌ ÇÊ¿äÇÏ´Ù. À̸¦ À§ÇØ º» ¿¬±¸¿¡¼­´Â ±â¾÷ÀÌ Á¦°øÇÏ´Â °ø½Ãº¸°í¼­¿¡¼­ ÁÖ°¡ »ó½Â ¹× Ç϶ôÀÇ ÀÌÀ¯¸¦ ¼³¸íÇÏ´Â ¹®Àå µ¥ÀÌÅ͸¦ ¼öÁý ÈÄ ºÐ·ùÇÑ´Ù. ÇØ´ç µ¥ÀÌÅ͸¦ ´ÙÇ× ºÐÆ÷ ³ªÀÌºê º£ÀÌÁî ºÐ·ù, ·ÎÁö½ºÆ½ ȸ±â, LSTM, ¾ç¹æÇâ LSTM, ÁÖÀÇ LSTM ¹× ÁÖÀÇ ¾ç¹æÇâ LSTM ¸ðµ¨¿¡ Àû¿ëÇÑ ÈÄ ¼º´ÉÀ» ºñ±³ÇÏ¿´´Ù. ¶ÇÇÑ, °¢ ÇнÀ¸ðµ¨ÀÇ ÀÇ»ç °áÁ¤ÀÌ Àΰ£ÀÇ ÀÌÀ¯ ¼³¸í ÆÇ´Ü°ú À¯»çÇÑÁö ÆÇ´ÜÇϱâ À§ÇØ ³ªÀÌºê º£ÀÌÁî ¸ðµ¨°ú SHAPÀ» ÀÌ¿ëÇÏ¿© ºÐ¼®ÇÏ¿´´Ù. ±× °á°ú, ÁÖÀÇ ¾ç¹æÇâ LSTM ¸ðµ¨ÀÌ Àΰ£ÀÇ ÆÇ´Ü°ú °¡Àå À¯»çÇÑ °ÍÀ¸·Î ³ªÅ¸³µ´Ù.
¿µ¹®³»¿ë
(English Abstract)
Recently, automation in investment has become a reality because of the development of AI technology. However, it has become more difficult for investors to understand the stock price prediction made by AI and decide whether to follow its prediction. Thus, it is necessary to provide a reason for the stock price forecast so that investors can understand it. To solve this problem, this study collects and classifies sentence data that explain the rationale for the rise and fall of the stock prices. We compare the cause classification performance of the Multinomial Naive Bayes classification, logistic regression, LSTM, bidirectional LSTM, attentional LSTM and attentional bidirectional LSTM models on our data. Additionally, we analyze the process of prediction using the Naive Bayes model and SHAP. Those analysis methods enable us to determine if the decision of each learning model is similar to human decisions on causes. As a result, the attentional bidirectional LSTM model was found to be the most similar to human judgment
Å°¿öµå(Keyword) ÅؽºÆ® ºÐ·ù   ¼¨Çø®   SHAP   ÁÖ°¡   ¿øÀΠ  ¼³¸í   text classification   Shapley   SHAP   stock price cause   explanation  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå